David Widerker independently developed a criticism of Harry Frankfurt's attacks on his Principle of Alternate Possibilities (PAP) that was similar to but stronger than Robert Kane's criticism in the 1970's. Frankfurt's Principle of Alternate Possibilities states that:
(PAP) A person is morally responsible for what he has done only if she could have done otherwise, that is, only if she could have avoided performing it.Frankfurt's attack depends on a very clever demon or intervener who has the power to block certain of the agent's alternative possibilities, while allowing the agent to act on that one of the possibilities that the demon wants to occur. Frankfurt then argues that if the agent chooses on her own that possibility, then the readiness of the intervener to prevent other possibilities is irrelevant to her decision.
(IRR) There may be circumstances that in no way bring it about that a person performs a certain action; nevertheless, those very circumstances make it impossible for him to avoid performing that action.In order to know whether to block the action of an agent, Kane and Widerker pointed out that there would need to exist a prior sign that the action was now going to occur. Such a necessary and sufficient prior sign would exist only in a deterministic and strictly causal universe. Since Frankfurt's intervening demon, much like Laplace's demon, has no way of knowing the actual information about future events - such as an agent's decisions - until that information comes into existence, such demons are not possible and Frankfurt-style thought experiments, entertaining as they are, can not in fact establish the compatibilist version of free will. Other philosophers, notably Alfred Mele and David Robb, Derk Pereboom, and others have developed more clever attacks on PAP. Widerker and others have developed further defenses, and there have been a number of such rounds. In his latest paper in the Journal of Philosophy (v 103, no 4, April 2006 , p.163), Widerker has said that he too can imagine conceptually possible situations where the agent’s decision is not causally determined, yet it is not within the agent’s power to refrain from it.
The avoidability-independent account of moral blame to which I have been calling attention here places me in the camp of those called "Source Incompatibilists" who hold that, although moral responsibility is incompatible with determinism, moral responsibility does not require that the agent could have avoided acting as he did. On this view, moral responsibility just requires that the agent was the ultimate originator of his act, that is, he performed the act without being in any way caused or nomically determined to perform it. Perhaps a better term for such a position would be “Frankfurt-friendly libertarianism.”Note that the discussion is less about whether the agent has libertarian free will than it is focused on moral responsibility. Widerker says:
I do acknowledge the intuitive force of Frankfurt’s claim that the agent in an IRR-situation is blameworthy for what he did, even though that he could not have done otherwise. But here, too, my way of accounting for this intuition differs from Frankfurt’s. Frankfurt sees the reason for the falsity of PAP in the fact that the unavoidability of the agent’s act (the factor that renders it unavoidable) did not play any role in the causal explanation of the act. I, on the other, hand, attribute the failure of PAP to the fact that the agent, in acting as he did in an IRR-situation, showed a lack of respect for morality (that is, chose on his own to act contrary to the requirements of morality without being morally justified in doing so), and to the observation that this negative verdict would still apply even if one were to assume that the agent’s act was avoidable. However, the difficulties I find with Frankfurt’s position in no way detract from the importance of Frankfurt’s achievement, that is, that by drawing our attention to the possibility of IRR-scenarios and their moral implications, Frankfurt has profoundly deepened our understanding of the notions of freedom and responsibility.
Excerpts from Blameworthiness and Frankfurt's Argument Against the Principle of Alternative Possibilities
From Moral Responsibility and Alternative Possibilities, ed. David Widerker and Michael McKenna, 2006, p.53-54
1. Introduction A widely accepted moral principle affirms thatPAP: An agent is morally responsible for performing a given act A only if he could have avoided performing it.In his seminal article 'Alternate Possibilities and Moral Responsibility,' Harry Frankfurt (1969) has attacked this principle. Central to his argument is the following assumption:IRR: There may be circumstances in which a person performs some action which, although they make it impossible for him to avoid performing that action, they in no way bring it about that he performs it. (Frankfurt, 1969, pp. 830, 837)Frankfurt contends that in a situation of the sort described in IRR (call it an `IRR-situation') the agent is morally responsible for what he did even though he could not have avoided so acting. Hence, according to him, PAP is false. To establish IRR, Frankfurt appeals to an example of the following sort:Jones is deliberating whether or not to keep a certain promise he made to his uncle. Unbeknownst to Jones, there is another person, Black, who for some reason does not want Jones to keep his promise. Black has the power and the means to force Jones to break the promise. But wishing to avoid showing his hand unnecessarily, he has made up his mind to intervene if and only if Jones does not show a sign of going to decide to break the promise. Call that sign 'S1'. If Jones does show that sign, then Black does nothing, knowing that in this case Jones will act as he (Black) wants him to act. (it is assumed that Black knows Jones very well in this regard.) Finally, suppose that Jones decides to break the promise for reasons of his own.It seems that in this situation Jones acts freely without Black's intervention, and therefore is morally responsible for what does. But, given Black's presence, it would appear that he cannot avoid deciding to break the promise. Call this example 'Promise Breaking.' In Widerker (1995a), I argued, adopting a libertarian position with regard to freedom and moral responsibility, that examples such as Promise-Breaking fail to establish IRR and that, therefore, Frankfurt's argument against PAP does not succeeds Most importantly, I tried to show that his argument fails when applied to simple mental acts such as deciding, choosing, undertaking, forming an intention, that is, acts which for the libertarian constitute the loci of moral responsibility. Essentially, my argument was this: either the sign S1, which Black employs as a sign for not intervening, is (or is indicative of) something that, in the circumstances, is causally sufficient for Jones's decision at t to break his promise or it is not. If it is, then the example does not describe an IRR-situation, since the latter requires that the decision must not be causally determined. On the other hand, if S1 is not so associated with Jones's decision to break the promise, if S1 is merely a reliable indicator of it, then there is no reason to think that Jones's decision was unavoidable. In either case, the truth of IRR has not been established. In this chapter, I wish to advance the debate over the strength of Frankfurt's attack on PAP in two ways. First, I shall consider some new attempts to provide examples of IRR-situations and contend that they do not succeed either. Second, and more important, granting (for the sake of discussion) Frankfurt's assumption that IRR-situations are possible, I shall provide a comprehensive defense of PAP for one sense of 'morally responsible,' that of moral blameworthiness.An agent is morally blameworthy for performing a given act A only if he could have avoided performing it.(Henceforth, whenever I refer to PAP, I shall be referring to this construal of it.) I shall develop my defense of PAP (i) by questioning the moral assumptions employed by Frankfurt in his argument against PAP; (ii) by articulating and defending the basic intuition behind PAP; and (iii) by indicating how a defender of PAP might go about explaining away the intuition many share that in an IRR-situation in which the agent performed a morally wrong act, he is morally blameworthy for what he did.
4. A Defense of PAP (the W-defense) (p.63-64) Can the defender of PAP go further? Can he also show that there are good reasons to accept PAP, reasons that would explain why in an IRR-situation the agent should not be deemed morally blameworthy for the act he performed? This question is important, since although many will agree that Frankfurt's argument against PAP is unconvincing, they may still think that IRR-situations provide intuitive counterexamples to PAP. They may emphasize the fact that in situations of this sort the agent acts freely, doing what he does for reasons of his own without being coerced or causally determined to so act. They may, therefore, rightly wonder why (in the absence of any excusing factor) the proponent of PAP still thinks that such an agent should be exonerated. Can the PAP-defender allay this worry? I believe he can. To show this, let us focus on a specific example of an IRR-situation, say, some (suitably revised) version of the Promise-Breaking example mentioned earlier. Now consider the following reply by the defender of PAP.
Let me grant, for the sake of discussion, that in the IRR-situation under consideration, Jones acted freely in the sense that what he did he did for reasons of his own without being causally determined or coerced to so act. Still, since you, Frankfurt, wish to hold him blameworthy for his decision to break his promise, tell me what, in your opinion, should he have done instead? Now, you cannot claim that he should not have decided to break the promise, since this was something that was not in Jones's power to do. Hence, I do not see how you can hold Jones blameworthy for his decision to break the promise.Call this defense the 'What-should-he-have-done defense' or for short the 'W-defense'. The W-defense points to an important reason why it would be unreasonable to regard an agent morally blameworthy in an IRR-situation. When we consider someone morally blameworthy for a certain act, we do so because we believe that morally speaking he should not have done what he did. This belief is essential to our moral disapproval of his behavior. Sometimes, however, such a belief may be unreasonable, for example, in a situation in which it is clear to us that the agent could not have avoided acting as he did. To expect in that situation that the agent should not have done what he did is to expect him to have done the impossible. By implication, considering him blameworthy because he has not fulfilled this unreasonable expectation would be unreasonable. The W-defense thus suggests the following general constraint on ascriptions of moral blame: PAE: An agent S is morally blameworthy for doing A only if under the circumstances it would be morally reasonable to expect S not to have done A. This principle, which may be called 'the principle of alternative expectations' (PAE), enables the PAP-defender to formulate an intuitive argument for PAP.
(1) An agent S is morally blameworthy for doing A only if under the circumstances it would be morally reasonable to expect S not to have done A. (2) If S could not have avoided doing A, then on pain of expecting him to have done the impossible, it would be morally unreasonable to expect him not to have done A. (3) Hence, if S could not have avoided doing A, then S is not morally blameworthy for doing A.28Note that PAE is a more general principle than PAP, since it can be used to explain why we sometimes exonerate an agent in situations in which his wrongful behavior was avoidable. These, for example, may be situations in which that behavior resulted from his being unaware of the causal consequences of some act of his, or where it can be traced to insufficient moral knowledge on his part. In each of these cases, there is no reason to assume that the agent could not have avoided acting wrongly. Finally, note that PAE can be further generalized so as to yield a full necessary and sufficient condition for ascriptions of moral blame.
A person is morally blameworthy for doing A iff in doing A he acted in a morally wrong way, and under the circumstances it was morally reasonable to expect him not to have done A.
8. Conclusion (p.69) There is a widely held view among moral philosophers that Frankfurt's attack on PAP constitutes a definitive refutation of the traditional libertarian conception of freedom and moral responsibility. Elsewhere, I have argued against this conclusion by questioning the coherence of the type of scenario on which Frankfurt bases his attack (Widerker, 1995a, 1995b). In this chapter, I have tried to advance the debate over the plausibility of PAP further (i) by considering and rejecting more sophisticated counterexamples to PAP, and (ii) by casting doubt upon the moral assumptions underlying Frankfurt's attack. Furthermore, I have tried to articulate the basic intuition underlying PAP, and have also indicated how the defender of this principle might go about explaining away the intuition many share that, provided Frankfurt-type scenarios are coherent, the agent in those scenarios is blameworthy for what he does. What I hope to have shown is that the dispute between Frankfurt and PAP-defenders is not merely a dispute over the question whether such scenarios are possible, but is a more fundamental dispute that rests on conflicting intuitions about our notion of moral blame.An Exchange between Widerker and Fischer